library(dplyr)
library(ggplot2)
my_data <- read.csv(paste0("https://raw.githubusercontent.com/owid/owid-datasets/refs/heads/master/datasets/",
"Mental%20health%20services%20across%20incomes%20-%20Wang%20et%20al.%20(2007)/",
"Mental%20health%20services%20across%20incomes%20-%20Wang%20et%20al.%20(2007).csv"))Homework 4
\(\\\)
Introduction
Questions I will look at are:
Are individuals with severe mental health conditions more likely to receive treatment than those with mild or moderate conditions?
Which country provides the highest percentage of mental health treatment?
Does a country’s healthcare spending (as a share of GDP) influence the percentage of people with severe mental health conditions who receive treatment?
Summary of Dataset: I will analyze a dataset from Our World in Data that examines the percentage of individuals who received any mental health treatment in the past 12 months, based on household surveys conducted in a sample of countries in 2003. My visualization breaks down the data by severity of mental illness to show the differences in the treatment received: severe, moderate, or mild. The “severe” category includes individuals diagnosed with bipolar I disorder, those experiencing physical symptoms from substance dependence (e.g., withdrawal), people who have attempted suicide and have a co-occurring mental health disorder, and anyone with severe functional impairment in at least two areas of life or a Global Assessment of Functioning (GAF) score of 50 or below. “Moderate” applies to someone who is addicted to a substance without experiencing physical symptoms of dependence, or those who report a moderate interference in at least one area of functioning due to a mental health disorder. “Mild” refers to any other disorders that don’t qualify as severe or moderate. The dataset also includes each country’s healthcare spending as a share of GDP, which provides useful context when comparing access to treatment across income levels.
Reliability: This data is reliable because it is based on standardized household surveys conducted by trained professionals across many countries. This data was also published in a peer-reviewed study and hosted by Our World in Data, a reputable source for datasets.
\(\\\)
Loading and tranforming the data
Let’s now load the data and transform it so that it is in a shape that is easy to analyze and visualize.
\(\\\)
\(\\\)
\(\\\)
Results
Analysis 1: Treatment Percentages by Country and Severity Level
\(\\\)
To start our analysis, let’s examine differences in treatment received across countries.
\(\\\)
\(\\\)
\(\\\)
Analysis 2: Country’s GDP on Healthcare effect on Severe Treatment Percent Received
\(\\\)
To figure out if a country’s GDP affects the percentage of severe treatment percent received, let’s look at an interactive plot!
\(\\\)
\(\\\)
\(\\\)
Conclusion
In this analysis, we explored the relationship between severe mental health treatment and countries. We found that:
Yes, people with severe mental health conditions are more likely to receive care than those with moderate or mild conditions.
The U.S. provides people with the most mental health treatment.
While the U.S.’s high spending on health coverage leads to a higher mental health treatment coverage for those with severe mental health conditions, this trend is not universal. The correlation value of 0.52 suggests only a moderate relationship, meaning other factors can influence access to treatment.
Remaining Questions: Questions that remain are why are those with severe mental health conditions more likely to receive care than those showing moderate or mild conditions, and how do they change over time. Since this data is from 2003, I wonder whether the trends would remain consistent in more recent years.
\(\\\)
Reflection
I chose to revisit the dataset I first worked with earlier in the class, and it turned out to be especially well-suited for this assignment. It was easy to transform my data into a data frame easier to read. Also, adding stickies, narratives, and highlights was something I found fun doing. Creating some of the visualizations was challenging, particularly the first bar plot, where I had to carefully align three bars per category to represent the different severity levels. It was also pretty difficult getting the perfect values when zooming in. I had initially tried to incorporate a visualization with multiple line plots, each of which would correspond to a country, but I ended up using a barplot instead to incorporate the different types of severity treatment received. In total, I spent more than 10 hours on this project.